Enter opens the highlighted result. Esc closes.

Qwen3.6 27B vs Qwen3.6 35B A3B

Qwen3.6 27B leads Qwen3.6 35B A3B on the Artificial Analysis Intelligence Index, 21.4 to 18.2, and Qwen3.6 35B A3B costs 2.7x less per output token ($1.00 vs $2.70 per million). Updated October 23, 2026; 17 benchmarks have a result for both.

Replace Qwen3.6 27B Replace Qwen3.6 35B A3B
Qwen3.6 27BQwen
Coding
36.8
Overall
45.3
Output $/M
$2.70
14
Qwen3.6 35B A3BQwen
Coding
No data
Overall
45.0
Output $/M
$1.00
2
vs boards won of 17 shared, 1 tied

Qwen3.6 27B or Qwen3.6 35B A3B: which to pick

Only Qwen3.6 27B has a coding score, 36.8 on the Vibe Coding Index; Qwen3.6 35B A3B has none, so the two cannot be ranked on coding.

Pick Qwen3.6 35B A3B for output tokens 2.7x cheaper ($1.00 vs $2.70 per million) and work measured by τ²-Bench Telecom or Mystery game puzzles, where it leads.

Model AQwen3.6 27BQwen36.8/1002 independent benchmarks$1.30 per solved taskModel BQwen3.6 35B A3BQwenNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
17
Qwen3.6 27B only
8
Qwen3.6 35B A3B only
3
Like-for-like categories
6 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Qwen3.6 27B 36.8 from 2 benchmarks, Qwen3.6 35B A3B not benchmarked.
Widest category
Math, 9.6 points apart on average over 2 shared benchmarks.
Widest single gap
Mystery game puzzles: Qwen3.6 27B 7.0% against Qwen3.6 35B A3B 22.0%, 15.0 points apart on a 0 to 100 scale.
Output price
Qwen3.6 27B $2.70 against Qwen3.6 35B A3B $1.00 per million output tokens, 2.7x apart.
Context window
Both take 262K tokens.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryQwen3.6 27BBarsQwen3.6 35B A3BTies
Coding201 tie
Intelligence20
Agentic work11
Reasoning31
Math20
General40
All shared boards1421 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index20
Coding benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
AA Coding Index53.741.9Qwen3.6 27B
SWE-bench Verified (Vals AI)70.0%Not measuredOnly Qwen3.6 27B measured
Vibe Code Bench (Vals AI)11.9%Not measuredOnly Qwen3.6 27B measured
SciCode42.8%36.6%Qwen3.6 27B
Terminal-Bench Hard (Artificial Analysis run)34.8%34.8%Tie
Terminal-BenchNot measured24.6%Only Qwen3.6 35B A3B measured
WeirdMLNot measured34.5%Only Qwen3.6 35B A3B measured
Neither model has a result on the other 36 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
AA Intelligence Index21.418.2Qwen3.6 27B
Epoch Capabilities Index146.2143.7Qwen3.6 27B
Agentic workLike-for-likePublic boards, shown as published, kept out of the index11
Agentic work benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
AA Agentic Index18.513.1Qwen3.6 27B
τ²-Bench Telecom94.2%95.3%Qwen3.6 35B A3B
Neither model has a result on the other 40 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index31
Reasoning benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
GPQA Diamond (Artificial Analysis run)84.2%84.1%Qwen3.6 27B
Humanity's Last Exam (Artificial Analysis run)23.1%22.2%Qwen3.6 27B
GPQA Diamond (Epoch AI run)85.9%84.8%Qwen3.6 27B
Mystery game puzzles7.0%22.0%Qwen3.6 35B A3B
Neither model has a result on the other 19 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index20
Math benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
OTIS Mock AIME91.1%86.7%Qwen3.6 27B
FrontierMath, tiers 1 to 335.1%20.4%Qwen3.6 27B
Neither model has a result on the other 12 boards here
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
Knowledge benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
CaseLaw (Vals AI)53.2%Not measuredOnly Qwen3.6 27B measured
CorpFin (Vals AI)62.3%Not measuredOnly Qwen3.6 27B measured
TaxEval (Vals AI)71.3%Not measuredOnly Qwen3.6 27B measured
MortgageTax (Vals AI)68.3%Not measuredOnly Qwen3.6 27B measured
Neither model has a result on the other 16 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index40
General benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
SAGE (Vals AI)45.6%Not measuredOnly Qwen3.6 27B measured
IFBench67.6%64.4%Qwen3.6 27B
Long context reasoning (AA-LCR)77.3%71.7%Qwen3.6 27B
DTBench78.1%73.9%Qwen3.6 27B
LMCA34.5%29.7%Qwen3.6 27B
Surface Evolver BenchNot measured44.4%Only Qwen3.6 35B A3B measured
EQ-Bench 41026Not measuredOnly Qwen3.6 27B measured
Neither model has a result on the other 41 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Qwen3.6 27B and Qwen3.6 35B A3B
BenchmarkQwen3.6 27BQwen3.6 35B A3BLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Qwen3.6 27B$0.0018Fits in one request
Qwen3.6 35B A3B$0.0006Fits in one request

Qwen3.6 35B A3B costs less

Repository review

50K input + 3K output tokens

Qwen3.6 27B$0.031Fits in one request
Qwen3.6 35B A3B$0.0080Fits in one request

Qwen3.6 35B A3B costs less

Agent loop

200K input + 10K output tokens

Qwen3.6 27B$0.117Fits in one request
Qwen3.6 35B A3B$0.030Fits in one request

Qwen3.6 35B A3B costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Qwen3.6 27B and Qwen3.6 35B A3B
SpecificationQwen3.6 27BQwen3.6 35B A3B
MakerQwenQwen
WeightsOpen weightsOpen weights
ReleasedApril 27, 2026April 27, 2026
Context window262K262K
Input $/Mtok$0.45$0.10
Output $/Mtok$2.70$1.00
Coding index36.8Not benchmarked
Independent measurements behind it20
Cost per solved task$1.30Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Qwen3.6 27B better than Qwen3.6 35B A3B for coding?

Only Qwen3.6 27B has a coding score, 36.8 on the Vibe Coding Index; Qwen3.6 35B A3B has none, so the two cannot be ranked on coding. Pick Qwen3.6 35B A3B for output tokens 2.7x cheaper ($1.00 vs $2.70 per million) and work measured by τ²-Bench Telecom or Mystery game puzzles, where it leads.

Which is cheaper, Qwen3.6 27B or Qwen3.6 35B A3B?

Qwen3.6 27B costs $0.45 per million input tokens and $2.70 per million output tokens; Qwen3.6 35B A3B costs $0.10 and $1.00. Cost per solved task is known for Qwen3.6 27B only.

Which has the bigger context window, Qwen3.6 27B or Qwen3.6 35B A3B?

Neither: both take 262K tokens.

How many benchmarks is the Qwen3.6 27B vs Qwen3.6 35B A3B comparison based on?

17 benchmarks have a result for both models as of October 23, 2026. Qwen3.6 27B's coding index rests on 2 independent benchmarks and Qwen3.6 35B A3B's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next